Research says that 85 percent of people will make a purchase after reading online reviews about a product or service. This has had huge implications for the hotel industry and helps explain why TripAdvisor, a massive repository of user-generated reviews, was the most-visited travel website in the world in 2013. Professor Thales Teixeira discusses TripAdvisor’s staggering success, how the company has forced an entire industry to change the way it considers (and purposefully influences) the online review process, and how consumers navigate that sea of reviews.
Brian Kenny: Today we’ll hear from Professor Thales Teixeira about his case “Managing Online Reviews at TripAdvisor.” I’m your host Brian Kenny and you’re listening to Cold Call.
Professor Teixeira is an expert in advertising and digital marketing and the economics of attention. Thales, thanks for joining us.
Thales Teixeira: My pleasure.
BK: What motivated you to write the case? Why were you interested in it?
TT: Some of my research is on the economics of attention and online reviews have amassed a vast amount of attention nowadays. People have changed their behavior in terms of buying products and services. They spend much more time paying attention to reviews of other people like them and trying to understand and gauge in an unbiased fashion what the real quality is of these products and services, particularly those in which you can’t evaluate before you buy the product.
BK: You cite some statistics in the case about the trustworthiness of online reviews. It’s remarkable that eight 85 percent or so of people would make a purchase after reading online reviews.
TT: Yes. It turns out that if you think about it from a consumer standpoint: what are the available options of information that I have for trying to buy any service? I can basically rely on information provided by professionals and experts, and some people do that. I can rely on information provided by the company, and that’s advertising and PR and other things like that. Or I can rely on information provided by other consumers like me.
In the case of hotels, professionals tend to not be as well-regarded anymore. Why? Because they have different values of what they consider important than [the average consumer does]. If I have a family and [experts] start talking about how amazing the nightlife is in that city or how exquisite the hotel environment or bar is, I don’t care about that. What I need to hear is from somebody who has a family who stayed there and that’s where that 85 percent number comes from.
BK: And everybody has different lenses that they experience things through. So one person’s experience could be very different than another person’s experience.
TT: That’s when online reviews matter a lot—if you can get this variety of experiences, people with different preferences, people with different tastes, and people with different “value functions” as we call it. For example, people on business, people on honeymoon, and people with young kids all have very different expectations.
BK: So what started as an aggregation of consumer reviews is now blossoming into much more of a full-service type environment. Is there the potential there for some kind of bias to creep into that?
BK: You call the case “managing online reviews.” As I read it, I thought the notion of managing online reviews to me seems to be counter to the notion of providing unbiased information. Yet you do cite some examples in there of companies that are managing online reviews in a way that isn’t inauthentic. Can you describe that?
TT: Sure. It is natural to imagine that once any consumer touchpoint starts having an impact on people’s decisions, companies want to have an impact on that touchpoint. So with TripAdvisor it became the go-to destination for hotels and other travel services. What hotels started to realize is that if they don’t have good TripAdvisor reviews this really damages their business. It has become the most influential aspect. So studies done by Cornell’s hotel school showed that the biggest factor that explains peoples’ decision for hotels now is reviews, online reviews, TripAdvisor being the biggest. The second one is location of the hotel and the third is price. So companies control price. They don’t control location. They control where they set up but once they set up they can’t change that anymore.
So now what they’re trying to do is influence the first big thing, which is managing reviews. Now how do you manage reviews? There are three types of ways that you can manage reviews as a hotel. You can manage the location of the review. People are probably going to review your hotel. You can try to nudge them to review it at one particular place. For example, the Four Seasons tries to nudge you to write a review on their own website. Others try to ask you to write a review on TripAdvisor by putting these little signs up and asking you and sending you reminders—so the location of where you write such that they aggregate the reviews instead of becoming scattered all over the internet. The second one is the amount of reviews. The same way that you can manage the location of the reviews you can try to get people to review.
BK: More is better I assume.
TT: More is better to some extent. If you have very few reviews then each additional review will help. Now at the margin after you’ve had 100 reviews on TripAdvisor, what is the benefit of 101 reviews? Research was done to show that it’s very minor. You have to get scale at some point and then afterwards it doesn’t matter. Now the most important aspect of reviews are the quality of the reviews, how many stars you’re getting. And then we start to come into some trouble. How does a hotel manage the quality of a person’s review? There’s a study that was done by McKinsey and I put it in the case. McKinsey said, “We’ve studied many industries and the biggest impact, biggest factor that impact peoples’ decisions making is word-of-mouth.”By their claim they’re kind of saying companies should focus much less on advertising and much more on word-of-mouth. And so I challenged this in the case by telling the students, “Okay, so let’s just stop doing advertising, let’s just improve people’s word-of-mouth of our product or hotel.” .
BK: Which you do by providing a better experience for people.
TT: Yes, that’s the point. Word-of-mouth has always existed. If you have a better product, you’re going to have a better review. But that’s not what’s in this discussion. If you’re going to have a better product your advertising is going to work more. Apple’s advertisements work very, very well because they have amazing products.
The point here is not improving your product, because I could go to any hotel and say, “Oh you want to sell more, I know what you should do—have a better hotel, have a better experience, have better services.” And then word-of-mouth is going to be better and you’re going to sell more. And they say, “Well, that’s obvious.” Apart from the product, you cannot influence the quality of your reviews unless you start doing things that are very, very fuzzy. For example, some hotels when you check out they ask how your experience was. And if you said, “Oh it was amazing, I loved it, the staff are so good,” then they give you a form and ask you to please fill out a review on TripAdvisor and we’ll give you some…
BK: They’re manipulating you.
TT: Manipulating to some extent. You could argue what the definition of manipulation is. If it’s changing your mind, they’re not changing your mind because you already said that you love the experience. All they want from you is to incentivize you to talk about that experience. So it could be argued that that’s not manipulation. Now what happens with the person who says, “Oh, I had an awful experience, the staff was not courteous, my room was awful.” What are they going to do? They don’t ask you to write on TripAdvisor, so they withhold. They selectively try to choose the better reviews. And at that point in class the students start debating what is appropriate, what is not appropriate, what is actual manipulation, what is actually self-serving by selecting the better ones? So this idea that companies should start influencing word-of-mouth, marketers should influence word-of-mouth, is actually not entirely correct. We still do advertising because you control 100% of your advertising. You have very little, sometimes no control over what other people say about your product apart from the quality of the product itself.
BK: This is a lesson that actually applies to any consumer category really. I would say with pretty much any product or service that somebody is selling, word-of-mouth is always going to be important. So for the marketers that are out there listening to this podcast, is there one particular thing you would like them to take away from a case like this? .
TT: The particular thing is that while there’s been a lot of badmouthing of advertising, if done well it has an impact on your business. Now there are many ways to do it and some ways have better or worse impact or lower impact. But controlled, firm-driven communication using the brand, my definition of advertising, still has a place for many, many companies. They should track consumer-to-consumer communication, i.e. word-of-mouth, but in some cases with word-of-mouth there’s very little that you can do about it. .
BK: So that leads me to my last question, which is for the consumers that are out there listening to this. How concerned should they be about having to filter the authentic reviews from the inauthentic? Is it a rampant problem? Are many brands out there stacking the deck with inauthentic reviews? .
TT: There’s definitely some amount of influence that the industry is making on providing various levels of inauthentic reviews. When you talk to consumers, most of them say, “I’ve gotten pretty good at selecting which ones are inauthentic and just using the authentic ones.” And that’s one thing consumers say. The other thing is if there are enough reviews I can de-bias—I can just select the ones that are more authentic and not consider the inauthentic ones. .
BK: We use our trust meter that’s built in and figure out which ones are authentic. .
TT: Yes. And if you had a course in statistics, one of the simple things that you learn is if you’re sampling from a biased set, even if you have a thousand, a million, ten million samples, your sample is still going to be biased. There’s no amount that you can use to overcome a biased sample. Size will never eliminate the bias. Just knowing that is important, that just going to a website that says, “We have a million reviews,” might actually be worse for you to make the correct decision than one that has 10 or 15 reviews. What is important is how these reviews were collected. It’s very important when you’re looking at how the reviews are configured to understand who is allowed to review. And I think that would be my biggest suggestion: don’t think that more is better. In this case, it is not. .
BK: Very helpful as I think about planning my next family vacation. Thales, thank you for joining me today. .
TT: My pleasure Brian. Thank you for having me. .
BK: You can find the case “Managing Online Reviews at TripAdvisor” in the HBS case collection at hbr.org. I’m Brian Kenny and you’ve been listening to Cold Call, the official podcast of Harvard Business School.